Getting better signal out of Canon XL H1 or XH A1.
Hello, I am trying to make my Canon prosumer cameras last a little longer & am thinking about trying the following. I am curious if this signal path would result in a better final image.
Starting with my XL H1 - using the D-terminal to component output cable then running that through a Blackmagic analog to SDI converter, which would allow me to inbed the audio in the SDI signal, then output the SDI signal into a Blackmagic Hyperdeck Shuttle. I am trying to get to 4:2:2 colorspace and still keep the full HD video without the heavy compression of the internal HDV.
Please let me know if this will work or if I am just dreaming. Thanks for any comments.
Wow Pete... sounds like to me you are talking about REALLY jumping though a lot of hoops and running around the barn and back... only to get at best the same results as ingesting via firewire... or maybe even poorer.
I'm guessing you are talking about recording from your camera "live" right into your computer... not capturing footage you've already shot. Already-shot footage, recorded to the tape, will always be HDV at best, no matter what.
The only way to get full HD, prior to the HDV compression, is to shoot live and capture into your NLE using the H1's HD-SDI port. If you have one of the later-model H1 bodies, then the audio is embedded in HD-SDI. In the earlier H1 models, the audio is not embedded (I was never sure why Canon did the non-embedded audio... I think they pretty quickly learned that's not what people wanted so it was embedded in later versions of the camera).
Even shooting live, the component signal you are talking about using is after the HD-to-HDV compression. Plus, obviously, you are converting a digital signal to analog and then back into digital again. That analog step is the killer. The only before-compression signal is the one that comes out via HD-SDI.
All that being said, are you having trouble with something (keying or compositing) because you don't have the 4-2-2 colorspace? HDV has a lot of haters who will rag on it because of supposed compression issues... but it really is a darn good acquisition format. I'll note that it's a good format for acquisition only that is... you shouldn't stay in HDV land at all after shooting. When we shot with the H1 we always captured as uncompressed 10-bit AVI files (Matrox AVIs), so we were never in a compromised colorspace or compression levels any more. Before moving to the C300PL last year I shot with the H1 for about five years... and even with a lot of compositing and keying work I honestly never had even the tiniest problem with the H1's footage due to too much compression.
Fantastic Plastic Entertainment, Inc.
Thanks for the response. I realized after I posted the question that I was jumping through a lot of hoops to get a marginally better signal. I do have FCP 7 on my MacBook Pro laptop. The issues I am having is when I am trying to shoot green screen. I have a decent set-up & have really worked at keeping the background nice & flat. It just seems like I still get a little edging when I capture to tape & then ingest that footage from the camera.
I will try to capture straight to my laptop using the firewire to bypass the compression, as I do have one of the earlier versions of the H1 that does not have the embedded audio in the SDI. That is really a bummer & wish there was a way to upgrade that.
Thanks again, Peace Pete
[Peter MacNaughton] "I will try to capture straight to my laptop using the firewire to bypass the compression"
Sorry, no, even if you are shooting "live" (not capturing on tape) going via firewire does not bypass the HDV compression... firewire snags the signal after it has been compressed from HD to HDV in the camera. The only way to bypass the compression in the H1 is to capture via the HD-SDI output.
With your earlier-model H1 (which mine is also), the typical way to do that is to ingest video via HD-SDI while simultaneously capturing the audio via analog. Not all hardware will let you do that. In our particular case our suites all have Matrox machines (either AXIO LE or MX02), and those boards will let you take in HD-SDI and analog audio at the same time. But frankly we never did that back in our H1 days... we just captured via firewire.
I'm going to suggest that your keying issue is not so much with compression in your video, but with your editing software. The native keyers in most NLEs (FCP, Premiere, whatever) are not known for being that great. You'll have much better luck using something a little more robust for keying... we edit with Premiere (CS6), but usually click over to After Effects to do any greenscreen shots using KeyLight to do the chroma compositing. It's pretty flawless, and infinitely better than the keys we get next door in Premiere.
Fantastic Plastic Entertainment, Inc.
The Atomos Samurai recorder allows you to record from the HD-SDI directly into ProRes 422 (or Avid DNxHD) and will also accept an analog line in signal which it will encode simultaneously into the ProRes file. So for the XL H1, you would feed the video signal from HD-SDI into the recorder and take the audio from the RCA sockets into the line in.
After Effects CS6/ FCS3 / Canon XLH1 / Canon 7D / Reason / Cubase
"it's either binary or it's not"
Hey Declan -
Thanks for the info I really appreciate it. I have wondered if I could run the audio separately into something like the Samurai & if it would keep everything in sync. I will investigate that some more.
"the component signal you are talking about using is after the HD-to-HDV compression"
I couldn't dissagree more. I have two cameras and our tests gave clear evidence in many different way's of testing that component output on XHA1 is uncompressed.
Common sense was telling me that there is no point giving analog output of compressed signal. Canon cut corners on XHA1 camera when coupled new imaging tech with oldfashioned tape: The very reason I bought it and never regret it, is that records on tape. All HDD/FLASH camcorders of that time (2006) were either much more expensive or had much worse image than XHA1. My common sense was teling me, that CANON have boat loads of tape mechanisms on its disposal, so most of my money will go to image quality, while in new flash based cameras, much of cost went into recording tech. That proved many times on field when I was "fool" to still have tape, compared to my colegues, but in editing room, they were "fools" having much worse image :)
So why would CANON raise it's manufacturing costs by implementing another dedicated D/A converter for feature that 0,5% of users will ever be bothered to try?? It's 2014 and I "waited" that much time until I bought decent LCD with component-in, and Atomos and all stuff came from the sky to earth with asking prices.
I have my first camcorder for 7 years. I bought new second XHA1s in 2012 and 2 (used) HV40 few months ago. I really believe in this little HDV cameras for light stuff like lectures, churches, theathre plays... I have one DN-60 recorder, and wanted to couple other three cameras with it too. But since DN-60 is same money as Atomos, again my "component to ProRes" idea came along.
PS: HDV is OK codec if:
- you do not plan to alter it later
- grade it
- color correct it
- if camera was perfect still during shoot
- if frame does not include moving watter or foliage
- if not shooting formula one or football (fast paning)
HDV is like JPEG: A final, delivery codec. But in computer we can give 2-pass VBR encoding, to reasign bits to more demanding scenes. Camera compress CBR and makes big blocks in picture when moved fast. There is no blocks on component out, no banding, no moskito effect, ...
I can assure that XHA1/XHA1s's component output is NOT HDV compressed.
I have both cameras and I have DN-60 firewire recorder too. But I want to upgrade it to Atomos Ninja 2 or Blade, because of bad HDV codec. So I crawled the web and found this site http://teamjaded.com/2011/04/aja-ki-pro-mini-canon-xh-a1.html/comment-page-...
There I wrote long comment about my findings. The author of that blog comfirmed me the same, that they gained better color information for grading.
I'm certain that XHA1 gives uncompressed output via component and with less crippled colors and a bit sharper image too. But original grain that is lost due to HDV compression is now visible - how fine it depends on gain settings. It should be mentioned that XHA1 have lots of grain even at +3 dB, so only -3dB is relatively clean. But all that is not visible from badly compressed HDV.
The only crucial and weak link in that chain will be cheap Neet® - Component to HDMI Converter I plan to buy. But many users which tried it report that fine grain was introduced uppon conversion. I'm shure that grain is not introduced by A/D converter, but rather converter is so good that converts all the grain from camera. Bad converter would smear or blur grain!
That triggered my decision. Sometimes even bad review of product gives me right information. :)
How has the XH A1 been for you now that you are tapeless with it? I still use my XLH1, XLH1s and XHA1, but of course now want to go tapeless and I see that you have done that...
I am thinking of going...
Component to (Component to HDMI Converter) HDMI to Atomos Ninja 2 Monitor/Recorder
Component to (Component to HD/SDI Converter) to Odyssey7 Monitor/Recorder
Anyway, what have you come up with (converting) and has everything been worthwhile? Any insights would be worthwhile.
Unfortunately I did not finish my tapeless project yet with XHA1(s). I had unexpected life situation, so project stuck at some point.
I bought "Neet converter", but that does not work with Atomos. Atomos said it worked with original Ninja, but not with Ninja 2 any moore. While at Neet said, it must be digital copy protection in chip which prevents signal through.
It does fine if XHA1 is connected to LCDTV, though. Comparing component to LCDTV with component>Neet>hdmi>LCDTV, picture is identical, so Neet is good converter, but it didn't work. Same if connect BlurayPlayer with Atomos2.
So my illness set me back for half a year, so I did not order another C>HDMI converter (like Startech.com), to try it out.
What I do have and can confirm (since many debates over internet).
- I use Ninja2 with HV40 cameras (where benefits are great*)
- I can see benefit if XHA1 is connected via FireWire or via Component to TV**
- I do have XHA1 already tapeless with Datavideo DV-60***
So I hope I will finish my project soon. My goal is to have one XHA1s coupled with Atomos Ninja Blade for great low light camera with great monitor, other older XHA1 can stay with DN-60, than both HV40 with Atomos Ninja2 or even Ninja star (I must test Star when CFast will be somehow cheaper).
For shooting on controlled scenes, XHA1 lacks great monitor and great codec. Ninja Blade is my goal. I'm confident, that uncompressed signal will give great benefit over HDV, even more at lowlight where 3CCD have more grain than CMOS. That will yield more natural (less compressed picture with more detail that can be pulled out with Neat Video, where HDV destroys much of fine detail). In controlled environment extra devices are not "bulky".
But for running and Gunnig I must consider. Even DN-60 is BULK on camera. For that sole scenarios, great quality TAPE is OK for hour of shoting. Maybe Ninja STAR will be ok for that (for other XHA1).
On Canon HV40, Atomos ninja 2 makes BIG BULK. So it diminishes handiness of camcorder, but I did not purchase 2 of them for anything else than "B-role" shots which are fixed at some position while doing interviews, live performances, etc. HV40 is better anytime than Gopro, having all settings needed and good optic. For anything important I have XHA1s or GH3 Lumix.
When I do live stuff (like stage play), 4 cameras makes 8 tapes = at least 5 hour of simultaneous digesting and 100€ for tapes. That cost me some 150€ per project. In some 10 projects I will return my investment, plus pisture is better, plus I can always use Atomos with future cameras or sell it. Plus work is done faster.
* HV40 benefit some 10% in sharpness and detail recorder via HDM to DNxHD220 vs HDV
* HV40 benefit A LOT in color definition, where you can now really crank CCR
* HV40 benefit somehow in natural appearance because more "analog-like" grain (noise) is visible instead of digital blockiness. And this noise is moving "live" (refreshed at 25fps interval), since HDV compression have keyframes where noise is somehow present, than few b-frames where not present (while playing, noise or rather blockiness is "flashing" and "jumping" instead being constant). Neat video work better with constant and uniform noise than with HDV.
* HV40 benefit a lot where moving and complex scene like watter or bushes.
- HV40 is a lot bigger with Atomos, start/stop does NOT work from camcorder and HV40.
+ I'm happy with improvement!
> here's first video I tested with HV40+Atomos Ninja2
** XHA1 show similar benefit not being compressed as HV40 if connected to LCDTV. I made test scene in my studio. I lighten with hot spots and dark shadows than I put different stuff in front of camera including rotating colorfull umbrella. I recorded HDV via DN-60 which clones internal HDV via FireWire on CF card. DN-60 have playback feature, so I could than switch between HDV playback and live feed from XHA1 to my LCDTV (via component out).
** XHA1 benefit at being less compressed (no "mosquito" effect near contrasty edges, no blockiness at moving objects).
** XHA1 benefit with better an more accurate colors and better percieved sharpness because color sampling is not HDV crippled.
I tested with all image improvements and noise reductions off on TV. Than I tested with sharpening boosted and with "resolution+" swithced on, to emphasize noise and artefacts from camera. I'm confident, that end benefit will be worth it. Right now I can say if better codec resolve some resolution too, because my test were not very accurate, since switching between DN-60 and live feed takes some seconds. Blindly I checked with my spouse too.
*** DN-60 IS CRAP!! Now DN-60A is out, allegedly better and fixed device. But this is contrary to XHA1 itself, very much "last year snow". They never fixed pre-record with HDV and timelapse. FireWire tends to make trouble with small miniDV terminal on XHA1 (big chunky cable and so fragile port). On XHA1s I dare to walk and shoot, but not with my XHA1. Device is not 100% reliable (ok, neither is tape). Conforming files takes good 5-10 minutes for 32GB, than at least 10-15 minutes on card reader (USB 3.0 goes faster, but DN-60 does not take latest UDMA7 CF cards), and it works exclusively only with Sandisk CF (which are expensice and CF is absolete for anything else).
DN-60 is made from crappy plastic - it breaks with total care in mind, not at all field device. Awukward design by all means - firewire at bottom, battery compartment (lid broke), CF card compartment (lid broke), menus, knob... I just use it because I payed so much once. And it is same price new is Ninja Star. And DN-60 is good ONLY with SD/HDV... Atomos will live with new cameras too.
So... I hope I'm not so much long you'll describe it TLNR :)
If you're like me, you'll want to read it all and moore. I digged over net for minths, readed several times everything.
I can confirm: XHA1 have no compressed output from component terminal and is in many aspects near XF300 (I owned for short period in 2012, but than I rather bought another XHA1s instead and saved tons of money). It only stuck with HDV codec... But I just love CCD (software stabilisation just works wonders, not like CMOS).
++invest time into QTGMC - it's really worth it!! I shoot 50i than convert to 50p and results are just stunning. It's like new camera, almost double the percieved resolution, color is steady and smooth (no color "jumping" like with HDV in Premiere). I can't wait to feed DNxHD recorded signla to QTGMC.
Thread bump alert!
I have an old Canon XH-A1 camcorder and a new Atomos Ninja. It seems that when I use the HD component break out cable into a converter to HDMI and into the Ninja, I can only get 30fps to record. But I'd like to shoot the pretty XH-A1 image at 24fps. Is there some work around to make this happen?